A tooth-bearing maxilla referable to Lycorhinus angustidens Haughton, 1924 (Dinosauria, Ornithischia)

ANNALS OF THE SOUTH AFRICAN MUSEUM
ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume

99

Band

December

1990

Desember

Part

10

Deel

A TOOTH-BEARING MAXILLA REFERABLE
TO LYCORHINUS ANGUSTIDENS HAUGHTON,
1924 (DINOSAURIA, ORNITHISCHIA)

By

C. E. GOW

Cape Town

Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material
becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town 8000

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na gelang van die
beskikbaarheid van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad 8000

OUT OF PRINT/UIT DRUK

1, 2(1-3, 5-8), 3(1-2, 4-5, 8, t.-p.i.), 5(1-3, 5, 7-9),

6(1. t.-p.i.), 7(1-4), 8, 9(1-2, 7), 10(1-3), 11(1-2, 5, 7, t.-p.i.),
14(1-2), 15(4-5), 24(2), 27, 31(1-3), 32(5), 33,

36(2), 43(1), 45(1), 84(2)

Copyright enquiries to the South African Museum

Kopieregnavrae aan die Suid-Afrikaanse Museum

ISBN 0 86813 117 2

Printed in South Africa by
The Rustica Press, Pty., Ltd.,
Court Road, Wynberg, Cape

In Suid-Afrika gedruk deur
Die Rustica-pers, Edms., Bpk.,
Courtweg, Wynberg, Kaap

C973

A TOOTH-BEARING MAXILLA REFERABLE TO
LYCORHINUS ANGUSTIDENS HAUGHTON, 1924
(DINOSAURIA, ORNITHISCHIA)

By

C. E, Gow

Bernard Price Institute for Palaeontological Research, University
of the Witwatersrand, Johannesburg, South Africa

(With 7 figures)

[MS accepted 5 December 1989]

ABSTRACT

The taxonomic status of the genera Lycorhinus and Heterodontosaurus, and the various
specimens referred to them, is in an unsatisfactory state, due to incomplete preparation and
description of otherwise good material, a tendency to diagnose specimens rather than species,
and a lack of understanding of the anatomy and functioning of the teeth. A new tooth-bearing
maxilla is described in detail, and both it and the type and only specimen of Lanasaurus scalpri-
dens Gow, 1975, are referred to Lycorhinus angustidens Haughton, 1924, which is also
restudied and reinterpreted here. Although detailed descriptions of the dentition of Hetero¬
dontosaurus tucki Crompton & Charig, 1962, have not yet appeared, L. angustidens and
H. tucki are readily distinguishable on postcanine tooth morphology, angle of wear facets, and
pattern of occlusion. Authors have ranked these differently specialized contemporary species as
primitive and advanced; this practice may be questioned.

CONTENTS

PAGE

Introduction. 367

The new maxilla. 368

Lycorhinus angustidens . 371

Discussion. 379

Acknowledgements. 379

References. 380

INTRODUCTION

In 1984 James Kitching and the author collected a tooth-bearing left maxilla
of an Early Jurassic ornithischian dinosaur. Any new material of these rare and
incompletely known animals is to be welcomed. As the specimen was studied it
became apparent that it is a larger specimen of the species Lanasaurus scalpri-
dens Gow, 1975, and that both are referable to Lycorhinus angustidens
Haughton, 1924. Impressions were accordingly made of the type of L. angusti¬
dens, thus enabling a detailed study of the three specimens. The study shows
that Lycorhinus angustidens differs from Heterodontosaurus tucki Crompton &
Charig, 1962. Most described material resides comfortably in one or the other of

367

Ann. S. Afr. Mus. 99 (10), 1990: 367-380, 7 figs.

368

ANNALS OF THE SOUTH AFRICAN MUSEUM

these species, with the exception of Lycorhinus consors Thulborn, 1974. The
much discussed specimen UCL A100 (Thulborn 1970) was probably correctly
identified as L. angustidens; it cannot be grouped with L. consors as Abricto-
saurus Hopson, 1975. The intention of this paper is to refrain as far as possible
from discussing the work of previous authors but rather to concentrate on
presenting new facts and inferences.

The following abbreviations are used to indicate the repositories of the
material studied:

BP — Bernard Price Institute
SAM — South African Museum
UCL — University College, London.

THE NEW MAXILLA

Locality

The farm Bamboeskloof, Lady Grey: 30°45'S 27°12'E, map reference
Floukraal 3027CC.

This locality is less than 15 km from two Heterodontosaurus tucki localities
in the Herschel district (Crompton & Charig 1962; Santa Luca et al. 1976) and
approximately 130 km from the type locality (Mount Fletcher) of Lycorhinus
angustidens Haughton, 1924. Lanasaurus scalpridens Gow, 1975, was found
about 250 km to the north (Golden Gate Highlands National Park). (See
outcrop and locality map in Kitching & Raath 1984, fig. 1.)

Material

The specimen (BP/1/5253) had been exposed to the elements for some time
prior to collection, with the result that the more delicate dorsal and anterior pro¬
jections of the maxilla are missing. The cutting edges of the teeth are also
damaged and two crowns are missing.

Preparation

Only a little mechanical preparation was necessary. This was followed by
treatment with thioglycolic acid, but this was discontinued as some damage to
the specimen became evident; this was in any case only cosmetic preparation.
Useful X-ray plates were made from the specimen.

Description (Figs 1, 3-5)

In occlusal view (Fig. 5) three important features are seen: the pit for recep¬
tion of the lower canine, the deep cheek region, and the pronounced curvature
of the dental arcade. The dentition is fully developed and well worn, indicating
that this was a mature individual. There are 14 functional teeth and a rudimen¬
tary 15th. The teeth invite several descriptive analogies; they are broadest
linguolabially and closely packed like a row of kernels on a maize cob. (The

TOOTH-BEARING MAXILLA OF LYCORHINUS ANGUSTIDENS

369

teeth of Lanasaurus scalpridens are broadest mesiodistally, but this difference
can be attributed to age—compare Fig. 6.) Lingually and labially the crowns
stand out from the roots (as evident in the photographs and indicated by dotted
lines in Fig. 1). This swelling of the crowns is reciprocated by swellings of the
roots at the mesial and distal ‘gum lines’ (for example see tooth 5). The final
analogy is that the teeth have a symmetrical cold-chisel shape with a constant
included angle of about 75° between wear facets and labial crown surfaces. (The
narrower teeth of L. scalpridens have an included angle of 45°. Thus, although
the angle is age dependent, it is worth stressing for comparison with Heterodon-
tosaurus that the new maxilla belonged to a mature individual.) This included
angle is a useful means of comparison as it is not affected by damage to the
cutting edge of the crown and it eliminates subjective reference to vertical and
horizontal axes.

More anterior teeth have taller crowns, whereas mesiodistal crown width
increases towards the back of the tooth row. The crowns have mesial and distal
ridges on their lingual and labial surfaces separated by grooves from the main
body of the crown; the distal ridges and grooves are more pronounced; on the
lingual surface grooves persist for some time as wear proceeds and are thus
important for assessing ages of teeth.

In the following text teeth are referred to by numerals for convenience. Part
of a wear facet is preserved on 2 but this tooth and 3 have lost much of the
crown tips; 5 is lightly worn with part of the posterior groove still present; 6 is
younger than its neighbours, being very little worn; 7 appears to have complex
wear, but the two small basal facets were probably induced by trapped food
rather than direct tooth on tooth contact (this is an old tooth); 8 is also well
worn. Here one begins to see the pattern that persists from this point pos¬
teriorly, whereby adjacent wear facets on successive teeth were formed by a
lower tooth in staggered occlusion. This pattern becomes very clear when the
teeth are viewed normal to the wear facets (Fig. IB—lines on the left in the
figure separate inferred lower teeth). Steps between adjacent teeth in this occlu¬
sal view immediately show up young teeth and correlate with those teeth (in
Fig. 1C) that retain traces of a posterior groove (notably 6 and 9; 12 is more
worn). Tooth 9 is lightly worn, 10 is heavily worn with some blurring between
the two main facets, possibly the result of polishing by food but also possibly the
remnant of an earlier facet (as argued for Lycorhinus angustidens —see below).
Tooth 11 has a well-developed pair of wear facets. Tooth 12, though a moder¬
ately young tooth (presence of posterior groove), is complicated, as its anterior
facet is actually paired—the result of being opposed by two successive lower
teeth. This tooth also has a large food polish facet. Teeth 13 and 14 are well
worn (retention of the posterior groove on the latter possibly due to delayed
eruption of a suitable antagonist). The rudimentary tooth 15 indicates that this is
a fully elaborated, mature dentition.

Teeth 6, 9 and 12 form a series of increasing age and are clearly younger
than the two teeth that follow each. Replacement thus proceeds from back to

370

ANNALS OF THE SOUTH AFRICAN MUSEUM

A

B

c

5

6

7

8

9

10

11

12

13

1 cm

Fig. 1. Lycorhinus angustidens, BP/1/5253, left maxillary dentition. A. Labial view.
B. Viewed normal to the wear facets. C. Lingual view. Note: In A and B hatching denotes
broken areas. In C wear facets are hatched. Lines to the left of B indicated where lower teeth

met each other.

TOOTH-BEARING MAXILLA OF LYCORHINUS ANGUSTIDENS

371

front in the row. This is the same as the pattern described by Gow (1975) for
Lanasaurus scalpridens (Fig. 6) and similar to that by Hopson (1975) for Lyco-
rhinus angusticeps but with a slightly modified interpretation (see below).

This has interesting implications. In the described dentition, teeth within
triplets are arranged in order of increasing age from front to back, but after two
more replacements a stage would be reached when this order would be reversed,
and it was just such a stage that pertained in the maxilla that opposed the type
dentary of Lycorhinus, as demonstrated below. X-rays of the specimen reveal
root canals filled with dense haematite; these show that only a very thin layer of
maxillary bone roofs the deep tooth sockets: the canal fillings terminate at the
alveolar border. In the tooth sequence 6 to 13, root-canal fillings are present for
all except tooth 7—this is probably a quirk of preservation as 7 should be the
last in the series 13-10-7 to be replaced. X-rays of L. scalpridens reveal a full
complement of roots. Although these X-rays show no signs of replacement activ¬
ity, this is not a firm indication that replacement had ceased. X-raying is a non¬
destructive technique that should be routinely applied and improved.

Lycorhinus angustidens
Figs 2, 3, 7

The specimen, SAM-3606, has been well described by Hopson (1975, 1980)
but was re-examined for this study owing to the possibility (now considered
confirmed) that the new maxilla belongs to the same species. In order to take
impressions, the specimen was thoroughly wet and the excess water removed
with compressed air; a fabric-reinforced latex impression was then made, the
first layer being of a very watery consistency. Three impressions were taken and
all are equally good. The impressions were coated with a fine film of sublimating
ammonium chloride, and it is these that were studied and photographed.

The present interpretation differs slightly but significantly from that of
Hopson (1980). The first point, which has not been stressed previously, is the
marked curvature of the postcanine tooth row. When the canine is oriented with
its cutting edges in a sagittal plane, the postcanine row curves back strongly
labiad. (The new maxilla matches this curvature. The best way to see this is to
orientate the photographs of the occlusal view with the first three teeth in the
sagittal plane.)

The canine bears serrations on both edges (four per millimetre) as illus¬
trated by Hopson (1980, fig. 1). However, most of the distal edge of the tooth is
missing.

Postcanine 1 bears a small mesial cusplet and above it the margin of the
crown is damaged (i.e. there may have been other cusplets). The posterior half
of the labial surface of the crown is damaged and this looks like wear, as the
damaged area has a sharp but smooth enamel edge (the worn area is covered
with matrix grains firmly adhering to the dentine surface).

Postcanine 2 has a worn occlusal edge to the crown; it also has a mesiolabial
wear facet almost certainly formed when the erupted tooth made contact with

372

ANNALS OF THE SOUTH AFRICAN MUSEUM

the opposing upper tooth that had earlier been responsible for making the facet
on IM—the first indication of a staggered pattern of occlusion of upper and
lower cheek teeth, which is argued in detail later. The condition of the labio-
distal surface of 2 is not clear due to adherent matrix, but it does seem confluent
with the anterior facet on 3.

Postcanine 3 has a small distal cusplet high on the crown. There is some
conchoidal fracture of the dentine at the tip of the crown but this does not mask
two distinct wear facets dipping slightly away from each other. Tooth 4 is very
similar though more worn and better preserved. Hopson’s (1980) interpretation
of tooth 5, i.e. one major wear facet and a small polished area, is accepted.
Hopson interpreted tooth 6 in the same way but the larger lower facet is in fact
in perfect contiguity with the anterior facet on tooth 7, thus demonstrating the
presence of an upper tooth in overlapping occlusion with 6 and 7.

Postcanine 7 is an old tooth; its posterior wear facet bears a wide, deep,
smoothly rounded groove. This groove must have been formed by a step
between adjacent edges of occluding upper teeth at different stages of wear. A
small facet is present on the mesial edge of 8. By tilting the specimen it is poss¬
ible to see that this facet lies on the same arc as the distal facet on 7—these
facets are thus attributable to the same upper tooth. The author is not convinced
that there is sufficient evidence for the same situation pertaining between 8
and 9, but agrees with Hopson (1980) that it seems likely. The large wear facets
on 8 and 9 have deliberately been left unhatched in Figure 2 because these teeth

Fig. 2. Lycorhinus angustidens, SAM-3606. Impression of left dentary teeth. Hatching indi¬
cates wear facets but has been deliberately omitted from teeth 8, 9 and 10. Teeth 8 and 9 bear
striations, 9 and 10 have heels shown by shading, and 10 bears a raised ridge with the same
orientation as the striations on 8 and 9. Wear facet on canine is on the lingual surface of the

tooth.

TOOTH-BEARING MAXILLA OF LYCORHINUS ANGUSTIDENS

373

bear patches of striations indicating direction of bite. These striations are helpful
to understanding the bite, which is seen to have a posteriad component. It is
important to note that the striations have the same orientation as the ridge on
tooth 10 discussed below. That such striations are rare suggests a degree of
imprecision in the bite such that occluding surfaces are continuously roughly pol¬
ished. Teeth 9 and 10 have heels worn into the base of their facets. The present
interpretation of 10 and 11 differs from Hopson’s but is made with the benefit of
the hindsight afforded by the new maxilla. Tooth 10 has a raised ridge between
facets, such as would result if the edges of occluding uppers did not quite meet.

The preserved portion of the 11th tooth was clearly part of a perfectly
normal full-sized tooth; it is faceted and is raised labiad of 10, and it was thus
opposed by the successor to the tooth responsible for the posterior facet on 10.
It is suggested that the differences in wear facet orientation that Hopson (1980)
recorded (supposedly increasingly horizontal with age) are illusory, as witness
the continuity of facets on 7 with those of its neighbours. Indeed tooth 4 seem¬
ingly has the most nearly horizontal wear facets, but is less worn than tooth 5,
which apparently has more oblique facets. This specimen represents a mature
animal of a species characterized by very oblique wear facets. Three more teeth
could have been present in the living dentary (see Fig. 3).

This dentition contains ample evidence of a staggered occlusal arrangement
between upper and lower teeth. In Figure 2 vertical lines above teeth indicate
where upper teeth would meet each other. Some of the most interesting and
instructive lower teeth are those that at first sight apparently do not conform to
this staggered pattern. The best place to begin is with tooth 6: here it is seen
that a second wear facet has started to encroach on a previously existing single
facet—the new facet would continue to enlarge and migrate forward as indi¬
cated by the arrow. One can postulate that exactly the same thing would happen
in time to tooth 5. Teeth 8 and 9 differ in that they have very well-developed

Fig. 3. Lycorhinus angustidens. Composite drawing of BP/1/5253 and SAM-3606. The speci¬
mens fit rather well and give an indication of the degree of incompleteness of the dentary tooth
row. For reasons explained in the text, wear facets cannot be directly compared.

374

ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 4. Lycorhinus angustidens, BP/1/5253. Above. Labial view. Below. Lingual view.

Scale bar = 1 cm.

TOOTH-BEARING MAXILLA OF LYCORHINUS ANGUSTIDENS

375

Fig. 5. Lycorhinus angustidens, BP/1/5253. Above. Occlusal view. Below. View normal to

wear facets. Scale bar = 1 cm.

376

ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 6. Lycorhinus angustidens (Lanasaurus scalpridens), BP/1/4244. Above. Labial view.
Below. Lingual view. Scale bar = 1 cm.

TOOTH-BEARING MAXILLA OF LYCORHINUS ANGUSTIDENS

377

Fig. 7. Lycorhinus angustidens, SAM-3606. Original above. Positive impression below.

Scale bar = 1 cm.

378

ANNALS OF THE SOUTH AFRICAN MUSEUM

single facets, and incipient facets on their anterior edges. These latter facets
would migrate posteriad in time until these teeth reached the condition seen in
tooth 10, which has two distinct but very well-worn facets. After this the tooth
would be shed. The difference in direction of facet migration has little to do with
position in the tooth row, but seems rather to be related to the extent of wear of
the teeth. From the above, the following sequence can be inferred.

(a) A single wear facet forms fairly symmetrically over the labial surface of the
crown of a dentary tooth (as previously noted by Hopson such a facet is
concave—the opposing upper teeth would be well worn and would present a
convex surface). This is not apparent from Figure IB and perceptions of facet
curves change as the specimen is rotated about its longitudinal axis. At its best
development each of a pair of facets on adjacent teeth is concave, hence a con¬
vexity is formed where they meet. This is seen in the occlusal stereophotograph
(Fig. 5) between teeth 7 and 8, and 10 and 11.

(b) A second facet forms posteriorly when a new upper tooth comes into occlu¬
sion. This facet migrates forward and eventually dominates the crown as the
tooth anterior to it is shed (this facet extends further down the crown—quite
obviously this must be so).

(c) As a new tooth comes into occlusion in the anterior position, a third facet
forms, this time on the anterior edge of the tooth, and this migrates backwards
to result in the condition seen on tooth 10. At this stage the tooth would be
replaced.

This interpretation highlights, and is itself supported by, the pattern of trip¬
lets in the Lycorhinus jaw. Arranged from youngest to oldest, these are 5, 6
and 7, and 8, 9 and 10. This interpretation differs from that of Hopson (1975,
1980), who proposed the following triplets: 4, 5 and 6, and 7, 8 and 9. Turning
to the anterior teeth, it appears that 4 is more worn than 3, but both have two
facets, whereas 2 has a single (first wear stage) facet; thus these teeth conform to
the pattern of triplets proposed here.

We now have the interesting situation where both maxilla and dentary bear
triplets of teeth that consistently range in age from front to back. For the lower
jaw one can demonstrate that each triplet would require to be opposed by a
battery of teeth in which the reverse situation pertained. To do this we can look
at the hypothetical maxillary (M) triplet 6, 7 and 8 that occluded with dentary
(D) teeth 6, 7, 8 and 9 at the time the bearer of the Lycorhinus type died.
Tooth M6 was well ground in, M7 had only recently made contact with D8,
whereas M8 may just have made contact with D9. We can also look at D9, 10
and 11: the oldest tooth in the next maxillary series should be M9—that fits;
MIO should be mature but not as old as M9 and again this is borne out by the
wear facet on Dll. All this makes eminently good sense, as, if occluding teeth
were to erupt together, the amount of attrition would presumably be greater and
the teeth would wear faster.

The pattern of wear on the teeth in the new maxilla has been frozen at a
different stage in the cycle, which makes it look different and more difficult to